Our team was recently asked for advice on how to modify an existing assessment task in a biology unit. The reason? ChatGPT, a generative AI tool, could provide an acceptable output within minutes.
Before
The original assignment had students present an academic article for 10 minutes in class. After that, their classmates would write a blog post about the study that included the pros and cons of the methods used in the study. The students were given access to the original paper after the presentation.
This task aimed to do three things: (1) get students to be engaged listeners, (2) strengthen their skills in critiquing studies, and (3) improve their ability to explain discipline concepts and the efficacy of the scientific methods used in the study to both experts and non-experts.
After (suggested)
Here are some options we suggested:
1. Shorter presentations with instant analysis
We suggested that students could present for only 5-6 minutes, shorter than the original 10 minutes. Right after, the ‘audience’ would use the ‘saved’ 4-5 minutes to answer questions about the strengths and weaknesses of the presented study as well as the key takeaways in bullet points. This change means students get to practice and demonstrate critical thinking, and they don’t have to prepare as much outside of class.
From the technical side, students could either use an iLearn Q&A forum where peers’ answers only become visible once a student has submitted their response OR an online survey form could be used to collect responses before students leave the room.
Having students respond in near real-time and in-class will greatly reduce the opportunity to employ a generative AI tool in an unauthorised manner, while gaining insight into each student’s critical analysis skills.
2. Localising Blog Posts
The blog post part of the task had to stay due to curriculum constraints (there was no time to apply for changes for next semester), but we suggested making the blog post more localised. For example, students could write about how the topic from the study relates to flora on campus and include photos illustrating the local example.
By going hyper-local it will be harder for a generative AI tool to do all the work.
3. Developing AI literacy
Instead of prohibiting AI tools, we suggested encouraging students to use them to develop AI literacy skills.
For example, students could be invited to experiment with different prompts in generating different versions of their post. Along with writing the blog, part of the task could involve sharing their top tips for creating effective prompts relevant to the task.
By deliberately scaffolding the use of generative AI tools, students have an opportunity to develop important AI literacy skills that are relevant to the discipline.
Let’s have a chat
Do you have an assessment task that you’d like to discuss/brainstorm with us? Feel free to get in touch with us at professional.learning@mq.edu.au. We are just an email away!
See also, other posts in the Generative AI series.
Image credit: Image generated with Midjourney (text to image) AI tool by Olga Kozar
Share this: